Well, it's freezing down in Texas
All of the telephone lines are down.
This a short piece about energy provisioning, resiliency and the role institutions play in shaping those outcomes. I’m going to talk about how organized wholesale energy markets work, how that differs from the older model of regulated public service, and how all that connects to the widespread power outages for households in Texas and beyond. Let’s start by identifying the key institutions.
Primary Institutions: Organized markets, regulated utilities, and public service commissions.
The Energy Reliability Council of Texas (ERCOT) is currently unable to serve peak load given its resources, so it’s doing some load shedding. This means it’s turning power off to specific load centers. Since ERCOT is an Organized Wholesale Energy Market, let’s start there. Organized markets are a huge part of the energy provisioning game, but they are weird and relative newcomers. Organized markets like ERCOT emerged in the ‘90s after passage of the Energy Policy Act of 1992, which established a federal framework for deregulation of the industry. Like most big policy ideas of the era, the intuition was that by deregulating the industry we could reshape the market to allow entrance of competitors thereby making it cleaner, cheaper and more reliable.
The old model - the vertically integrated utility public service corporation - was subject to state regulation and offered the monopoly franchise for their troubles. The basic concept is that the state recognizes that the activities of investor-owned-utilities is, by the their nature, “affected with the public interest,” and therefore subject to control by regulatory commission. Such controls include review and acceptance (rejection) of rate-making procedures, investment plans, and ensuring that a utilities is capable of meeting its public obligation. In exchange, the utility is protected from competition and allowed a ‘reasonable’ rate of return. What is a reasonable rate of return? That depends upon the law - read Commons on that. Or just read Yngve Ramstad on Commons, which is easier.
The deregulatory move might be most pronounced in Texas (most of the state receives electricity from unregulated suppliers), but most areas are served via a mixed-mode of provisioning institutions. In the end, what matters is whether the lights turn on when you flick the switch. Ultimately, a load-serving entity like a utility is responsible for making sure that happens. To ensure that all utilities in a region are capable of doing this, we have grid reliability institutions like ERCOT that are charged with planning and coordinating. In California, this institution is the California Independent Service Operator (CAISO). The North American Electric Reliability Corporation (NERC) is a quasi-public institution that serves as bulk power and reliability regulatory agency, governing each sub-region (like ERCOT), drawing its statutory authority to act as a regulatory body from the Federal Power Act (215(e)) the Code of Federal Regulation (Title 18, Section 37.9). And to enforce the law that shapes the provisioning of energy, there are the state-level pubic utility regulatory commissions and the Federal Energy Regulatory Commission (FERC). After listing all of these active and deeply integrated set of regulatory bodies, authorities, and institutions, it seems silly to cast any part of this game as “deregulated.” It is silly, because it’s not true. It’s better to think of it as reregulation or just simply a change in market governance technique. Anyway, that’s the industry parlance so we use it here.
Organized Markets are Weird
When I talk to energy economists this statement produces raised eyebrows and in some cases an emotional response that looks like Figure 1.
Organized markets are weird because they animate a lot of administrative rule-making and ad hoc procedures to compel a distributed set of market participants to act as if a market for wholesale energy existed in Nature. This is silly because markets do not exist in Nature. Like Market Design Micro Guys, they are built by institutions.
In Organized Markets the Independent Service Operator basically attempts to act as a Walrasian auctioneer, by looking at load and generating supply forecasts, solving a system of equations for the next day based upon those data as well as supply and demand bids from participants. The Center for Veb Account Research Newsletter has a good post on Walrasian stuff you can read here. There are other markets in the real-time, but the basic idea is this:
Get weather forecasts
Get supply forecasts
Make sure generation database is up to date, including cost of service.
Take bids from suppliers and load serving entities
Calculate the least solution for the day ahead market. Award bids accordingly.
Do some true ups on capacity once you run your market computer program
This is a central planning function. Say it back: The most deregulated parts of the electric provisioning process are also subject to central planning. So, even the deregulated markets are subject to central planning. This is not new, but it’s revelatory for some and uncomfortable for a few.
Planning for what and for whom?
All economic activity is subject to some method of central planning, so we should begin by asking what the planning activity seeks to achieve. For Organized Markets in the wholesale energy space, the objective is to minimize cost subject to reliability constraints. OK, this is fine. But, undergirding this institutional practice is a broader objective to reproduce a field in which entrepreneurs, investors, commodity brokers, non-profits, coops and traditional utilities can all get a piece of the action if they choose. Accommodating the desire to act as if there is a free market in energy provisioning, introduces a set of uncertainties that were not present in the old regulated utility model. The largest uncertainty lies in whether or not these participants will supply enough capacity resources that are flexible enough to respond to peak demand for energy load. The planning institutions for Organized Markets have methods designed to achieve grid reliability, but the preferred approach is to use scarcity pricing to induce market participants to offer costly capacity resources on certain hours to balance supply and demand for power. Unlike other commodities, this balance must be managed in real-time to maintain a standard frequency that our technologies depend upon. The hope as that the presence of rare but very rich hourly prices for energy will be sufficient to induce investors to develop expensive peaking resources. This works most of the time, but sometimes it doesn’t. When it doesn’t work the consequences can be severe.
If the objective were flipped: minimize grid reliability events subject to usual constraints, including contracts, compacts, formal and informal working rules, then you one could conceive of completely different provisioning picture. Without changing any of the other planning methods, you would see scenarios in which load-serving entities are required to carry more reserve capacity on conventional resources, more distributed generation and storage resources, as well as upgrades on the local distribution system. In the case of ERCOT, with a planning lens that sought to maximize resiliency, you might see effort to establish more interconnections with neighboring reliability coordinating regions and an improved grid. The difference between this solution and the least cost solution currently in practice, is an increase in the budgetary costs of the provisioning plant. But, with all cost - benefit calculations there’s a implicit valuation of which costs and benefits matter for some people and which do not. What is the net present value of the cost of losing a loved one to a cold snap?
My view is that we’re always planning for some kind of optimality, but it’s not always clear whose wellbeing we’re trying to optimize. If we want to plan for maximum resiliency, then we ought to change working rules and norms that govern the planning. We need to recognize that utility service is a basic need and is a very public affair, in which case we should fund its provisioning with public money. Let the engineers calculate the matching algorithms and figure out how to make the power flow, but let’s keep the economists out of the room.